"Discretion in Hiring"
Mitchell Hoffman, Lisa B. Kahn, Danielle Li

The anonymous data provider provided access to proprietary data. We cannot share the data.

Master.do runs all the analyses, beginning with Step1. Before Step1, various data assembly, anonymization, cleaning, cutdown, and merging operations are performed to create the base data files. For example, some original identifiers are removed and replaced; we extract the recruiter with whom an applicant has the most interactions; and data are assembled on unemployment rates.

Beginning with the base data files, the code proceeds as follows:

Step1 cleans the data and defines variables.

Step2 creates applicant pools.

Step3 creates our analysis sample and prepares the data for regression analysis.

Step4 produces all results in the main text, as well as in Appendix C Tables and Figures.

Step5 produces the tables and figures in Appendix A.


